9:41
All India Cut-off Ranks
Competition Level:
Very High
(
in 25')
2023
5,500
SR2
2024
5,500
Special..
2024
5,500
SR3
You (4506)
Detailed Report
Get shortlisted courses, admission chances by selected branch, city, fees plus counselling info
Download PDF
The College Predictor rebuilt for 1.5M+ students
Redesigning the admissions experience. Prediction summaries, admission probability insights, and downloadable reports that make complex counselling decisions effortless. Shipped during peak season, reaching 250K+ premium purchases.
Visit Website
View Case Study
Historical Data Analysis
5+ years of cutoff trends
Smart Predictions
AI-powered probability analysis
Verified Data
Official sources only
My role
Sr. Product Designer (IC)
Discovery → 0→1 → Launch
I owned
Research · UX Strategy · Product Design
Team
1 PM · Engineering
Sole Product Designer
Status
LIVE
Web + Android app
Context & Problem
A high-stakes decision, made in a hurry, with the wrong map.
Every year, ~1.5 crore Indian students sit entrance exams. In the 7–10 day window after results, they must lock counselling choices on platforms like JoSAA. The order they list colleges is irreversible.
Careers360’s College Predictor answered "what are my options?" at scale—matching ranks, categories, and states against 5+ years of cutoff data.
But watching real students during counselling week, the gap was stark: hit with a wall of 1,000+ results, they’d scroll in anxiety and abandon the tab. The critical question—“Which do I choose, and in what order?”—was left entirely unsolved.
The tool delivered data. It didn't offer direction.
1.5cr
Aspirants / yr across exams we cover
7–10
Day window to lock choices, no undo
1,000+
Avg colleges returned
per query
The moment that defined the brief
It gave me 1,000+ colleges. My counselling form has 20 slots. I still don't know what to put in slot one.
Aarav Chauhan (JEE Main 2025, Rank 41K)
Usability Session · Day 2 of JoSAA
BEHAVIORAL BREAKDOWN
The drop-off wasn't in discovery. It was at the decision.
Traffic was never the issue—the predictor is one of our most-visited tools. Mapped from session walkthroughs, confidence broke down right at action: not at "got my results," but at "what do I do next?"
Why this was worth a redesign, not a tweak
Raw Data, Zero Context
Answers "What are my options?" but fails to show "What should I do?" offering no guidance.
Every Card Looks Equal
Without visual hierarchy, a top-tier dream college and a safety backup look identical.
High-Stakes Anxiety
Because counselling locks are permanent, ambiguity creates instant stress for students and parents.
9:41
Scroll Ends, Decision Doesn't
Students scrolled a flat list with no quick way to shortlist or compare—so the value loop broke mid-funnel.
The Paradox of Choice
Showing 1,000+ options with equal weight didn't empower students—it caused choice paralysis.
Jobs To Be Done
What students were really trying to accomplish
Research
Nearly 75% of students never reached their predicted colleges.
We combined behavioral analytics, funnel analysis, session recordings, and heatmaps to uncover where students dropped off - and why.
374,082
Students
analyzed
75.11%
Exited before prediction results
100+
Behavioral sessions reviewed
Funnels
Locating exactly where 75% of high-intent traffic abandons the flow.
Heatmap
High engagement at the input fields, but total blindness below the fold.
Recording Sessions
Revealing real-time choice hesitation, fatigue, and sudden exits.
Problem
Choice Overload
Students faced thousands of unranked colleges, making it difficult to identify realistic options and move forward
Problem
The Trust Gap
Predictions lacked transparency, prompting students to leave and validate results on other platforms.
Problem
Wrong Mental Model
Students chose branches, not just colleges, but the experience was built around institutions instead of outcomes.
Problem
Decision Friction
Too many equally important actions increased hesitation at the moment students were ready to decide.
Problem
Broken Momentum
The journey ended with predictions, offering no way to shortlist, save, or confidently continue their admission journey.
Problem
Shared Decisions
Students and parents made decisions together, yet the experience provided no shared view to build confidence.
From data to insights
Design strategy
The Design Hypotheses
From the observations above, I framed the redesign as a set of testable bets - each one an if we… then… that a specific design decision would later have to earn.
Rahul
Students get 1,000+ results and freeze — abundance reads as overwhelm.
If we lead with a chance verdict - Good / May Get / Tough -then students orient themselves before they ever scroll the list.
Rahul
A row of equal-weight CTAs re-creates the overwhelm in button form.
If we set one primary action (Shortlist) and defer the rest, then students take a clear next step instead of stalling.
Rahul
A bare probability isn't believed — students cross-check 2–3 other sites first.
If we show the math — category-adjusted rank + cut-off history — then the number becomes evidence they trust.
Rahul
The decision dies inside the tool — a shortlist that never leaves is a dead end.
If we put shortlist, report & counselling one tap from every row, then the decision survives the jump to JoSAA.
Rahul
Predictions are per-college, but a student's real choice is per-branch.
If we compute chance per course, not per institute, then the result matches how students actually decide.
Rahul
Students and parents disagree, and neither has shared ground to argue from.
If we give a credible, shareable comparison, then the choice feels joint and defensible - not a gamble.
Our Strategic Approach
Analyze the Current State
2
Current State of
Initiative/Product
Current State of Intent
Set the Next Goal
3
option goal
Product Initiative
Understand the Direction
1
Product Initiative
Company Vision & Strategic Intent
Execution
Choose Step of Product Process
Problem Exploration Solution Exploration Solution Optimisation
4
Strategy Creation & Development
How I explored the concepts
Three honest bets to close the decision gap
I diverged hard before converging. I prototyped all three and pressure-tested them through internal testing and iteration; here's what I learned and why one won.
Concept A
Concept B
Concept C
Shipped
Concept A
The Enhanced
Filter List
Focusing on high scannability. Kept a dense list view but introduced top-level chance filters to allow instant grouping.
From Passive Data Dump to Guided Decision-Making
Instead of forcing students to process flat lists on their own, the redesign surfaces historical trends and planning tools in context to drive decision momentum.
Before
(Legacy Pipeline)
After
(The Dynamic Decision Loop
Good/May/Tough
Shows chances instantly
Shortlisted
Saves for later
Important Notes
Highlights key rules
Goal-Based Filters
Speeds up decision-making
Layout Toggle
Swaps card/table view
Eligibility Rank Mapping
Matches real ranks
landing page
Understanding intent
before prediction
Instead of hiding input fields behind multiple steps, the core predictive form is embedded directly into the first fold.
By pairing these fields with clear, trust-building value pillars, this single workspace minimizes top-of-funnel drop-offs and drives immediate decision momentum.
9:41
www.careers360.com
Trusted by 10L+ Students
JEE Main College
Predictor 2026
Get precise college predictions based on previous year's cutoffs and trends
Historical Data Analysis
5+ years of cutoff trends
Smart Predictions
AI-powered probability analysis
Verified Data
Official sources only
Enter JEE Main Paper-1 Exam Details
Get personalized college recommendations in seconds!
Exam Given
JEE Main Paper-1
Interested Counselling
Joint Seat Allocation Authority (JoSAA Counselling)
Your Rank
Home State
Delhi
Caste Group
General
OBC-NCL
SC
ST
EWS
Gender
Male
Female
Person with Disability (PwD)
Disability Quota
College predictor report will be sent to you on:
Mobile Number
Predict My Colleges
Enter JEE Main Paper -1 Exam Details
Get personalized college recommendations in seconds!
Exam Given
JEE Main Paper-1
Interested Counselling
Joint Seat Allocation Authority (JoSAA Counselling)
Your Rank
Home State
Delhi
Caste Group
General
OBC-NCL
SC
ST
EWS
Gender
Male
Female
Person with Disability (PwD)
Disability Quota
College predictor report will be sent to you on:
Mobile Number
Predict My Colleges
Trusted by 10L+ Students
JEE Main College Predictor 2026
Get precise college predictions based on previous year's cutoffs and trends
Historical Data Analysis
5+ years of cutoff trends
Smart Predictions
AI-powered probability analysis
Verified Data
Official sources only
https://www.careers360.com
9:41
9:41
9:41
9:41
Simple Inputs: Keeps the form fast and effortless
Verified Sources: Shows real data sources upfront
Immediate Proof: Eases hesitation right at the button
Clear Outcome: Shifts copy from a search tool to a guide
10M+ Scale Metric: Instantly proves credibility
Personalized Cards: Highlights clear value before they click
Real Student Stories: Builds confidence at the final step
Catch-All CTA: Gives long-scrollers an easy next step
https://www.careers360.com
9:41
www.careers360.com
Zero
Friction
9:41
Total Due
₹943
Get My College List
order summary
By anchoring the checkout flow to explicit value reminders and real-time social proof, this workspace minimizes late-stage drop-offs and accelerates transaction velocity.
Value Affirmation, Seamless Intent
9:41
9:41
9:41
9:41
Instant confirmation and clear receipt details eliminate post-payment anxiety and send users directly to their results.
Simple, reassuring messaging explains what went wrong without causing panic and gives a clear one-tap path to retry.
Active system feedback keeps users calm during loading intervals, backed by quick access to expert support if needed.
Clear plan comparisons and live social proof make value obvious upfront, reducing hesitation right before checkout.
9:41
www.careers360.com
https://www.careers360.com
From Static Results to
Confident Choices
Showed real-time activity right next to the filter and shortlist tools to build instant trust at the exact moment they decide.
We turned the result page from a passive report into a decision engine. By showing students clear prediction chances alongside live activity from peers, we removed final-step doubt and made the next step obvious.
Broke down over 1,000 college choices into simple, color-coded odds (Good, May Get, Tough) so users immediately see where they stand.
Instant Clarity
Peer Reassurance
How I explored the concepts
9:41
9:41
Placed a quick 1:1 counseling booking module directly in the flow for users who feel stuck or overwhelmed by choices.
Built intuitive, multi-tag filtering by branch and admission odds, letting students narrow down 1,000+ choices in seconds.
Brought key data—like fees, placement CTCs, and seat availability—front and center so students can evaluate options without leaving the page.
Surfaced multi-year closing rank trends directly in a clean modal to help users benchmark their actual chances with precision.
Every element on the card is engineered to eliminate guesswork and guide students toward high-intent action.
Personalized Rank Mapping
Plots the user's exact rank directly against historical cut-offs for an intuitive visual comparison.
Multi-Year Context
Surfaces 3-year historical trends and competition levels in place to set realistic expectations.
Frictionless Actions
Pairs shortlisting with instant secondary pathways - like side-by-side comparison - to keep exploration moving forward.
Upfront Verdict
Leads with clear, color-coded likelihood tiers (Good, May Get, Tough) so students know where they stand at a glance.
UX Decisions That Reduce Uncertainty
trade-off decisions
The hardest constraint wasn't technical-it was alignment.
The PM and I clashed over one thing: how many calls-to-action the screen should push. Working through it is what made the final direction strong — so I'm putting it first.
My Position · Design
One clear action per state. Everything else stays secondary
Compare, Shortlist, View Report, Book Counselling, Talk to an Expert, Download Brochure — stacking them all at equal weight is just the overwhelm again, in button form. A screaming row of CTAs makes the real next step impossible to find.
Ashok Singh
Sr. Product Designer
PM's Position · Product
Surface all CTAs to capture immediate conversion & revenue
Several of those actions are how the product earns its keep. Hiding them risked engagement and revenue, and the PM rightly didn't want to ship a beautiful tool that under-delivered on the roadmap.
Karan Goel
Senior Product Manager
How I Resolved It
Not "how many CTAs," but "which one is primary—and when do the rest appear?"
I set a clear CTA hierarchy instead of cutting actions. One primary CTA—Shortlist-gets the solid coral button on every row because it moves students toward a decision. Compare and View Report become quiet inline links. Business-critical actions—counselling, talk-to-an-expert, and direct admissions—surface as contextual cards deeper in the flow, only when relevant.
OUTCOME & MEASUREMENT
What shipped, what moved, and how the design proved its value.
The redesigned Decision Engine shipped live across JEE Main, EAMCET, and KCET variants — reaching 10M+ students. Replacing flat lookup tables with structured verdict tiers eliminated key drop-off points and drove an immediate surge in product purchases within 20 days of rollout.
-35%
Drop-off Reduction
Recorded across the first 20 days of live release.
40%
Purchase Rate
Jumped from 69/1,359 (~5%) to 400/1,000 (40%).
10M+
Student Reach
Active users across JEE Main, EAMCET & KCET.
VALIDATED HYPOTHESES
Driving high-intent product purchases
400/1,000 users (40%) purchased the product post prediction - up from 69/1,359 baseline
Result
Takeaway
Leading with plain-language verdicts guides students straight to value-adding paid offerings.
NORTH STAR SIGNAL
9:41
Eliminating triage paralysis at scale
-35% drop in drop-offs recorded across the initial 20-day launch window.
Result
Takeaway
Grouping scattered results into structured chance bands eliminated choice overload at the top of the funnel.
Friction reduction
9:41
9:41
Preserving algorithmic trust across edge cases.
10M+ students served smoothly across 14 distinct edge-case result states.
Result
Takeaway
Keeping data transparent enough that students & parents felt confident anchoring decisions and purchases on the output.
STRATEGY & TRUST
What worked
Live Season Research
Conducting research live during peak counseling season gave us raw, unfiltered behavioral signals.
Open Access & Trust
Keeping every college accessible via probability bands (Good, May Get, Tough) built instant trust—proving that transparency drives product purchases better than artificial gating
What I'd do differently & next
Post-Counseling Follow-Up
Follow up post-counseling to track if students secured seats in their predicted bands, feeding actual allotment data back into model precision.
Side-by-Side View
Evolve single-row predictions into a side-by-side comparison matrix so students can easily evaluate competing "Good Chance" choices.
What I'd carry forward
9:41
Decision Clarity
Data Overload
Guided Action
The win wasn't a prettier results page. It was a reframe - from "show options" to "support a decision"
The Real Breakthrough
Transforming 1,000+ raw cutoff rows into plain-language verdict tiers to eliminate choice paralysis at scale.
Find My Best Fit College